Abstract Voluntary refusal of recommended childhood vaccines is becoming more common. Moreover, spatial clustering of unvac- cinated children likely introduces epidemic outbreak risk that extends far beyond what population averages suggest. How- ever, the degree to which clustering of vaccine refusal affects risk of infectious disease outbreak is largely unknown. Dy- namic processes leading to clustering of vaccine refusal are also understudied, limiting our ability to target interventions explicitly at the high-risk phenomenon of vaccine refusal clustering. Many outbreak risk management strategies are based in policy (e.g., school-entry vaccine mandates). Such policies should consider the realities of clustered vaccine refusals. This K01 Career Development Award proposal consists of advanced training and research aims on this topic. The candi- date's career objective is to become a leader in infectious disease prevention by conducting research at the intersection of infectious disease, social epidemiology, geospatial science, and complex systems modeling. The training aims focus on three areas: 1) agent-based modeling of infectious disease, 2) health policy, and 3) professional development. An interdis- ciplinary team of four senior faculty mentors at UW-Madison and Emory University will oversee the candidate's training, and includes experts in clinical pediatric infectious disease, vaccine advocacy, vaccine refusal clustering, health policy, and agent-based modeling of health behaviors. The project has two research aims. Aim 1 is to quantify the excess risk of measles and pertussis outbreak stemming from clusters of unvaccinated children. Agent-based models will be used to sim- ulate populations and test dynamic within-group processes and potential policy effects on those processes. Aim 2 is to characterize the attributes of vaccine-hesitant ideas that promote uptake of hesitancy and refusals in terms of diffusion of innovation theory. A mixed methods approach will include structured interviews, a web-based survey of vaccine-hesitant parents (target n=600), and development of an agent-based model to aid in targeting future intervention efforts. Associa- tions between diffusion of innovation constructs in vaccine hesitancy and vaccine outcomes will be estimated. Study out- comes include: a) an estimate of outbreak risk stemming from clustering in vaccine refusal; b) a decision making tool to guide policy development; and c) characterization of vaccine hesitancy using a diffusion of innovation framework. These results will fill critical empirical and theoretical gaps in vaccine hesitancy by informing how to most effectively prevent the formation of (and how to respond to) clusters of parents who refuse vaccines. To maximize success and subsequent dissemination and public health impact of the proposed research, collaborators for the proposed K01 include experts in social networks, diffusion of innovation, agent-based modeling, and community-based early childhood intervention trials. With this network of collaborators, results of the mentored K01 research will be used to substantiate future R01 interven- tion trials using a preventive health framework to modify the dynamic group processes leading to vaccine refusal clusters and infectious disease outbreak.